Skip to Main Content
Beamforming techniques offer array gain and spatial multiplexing gain in multiple input multiple output (MIMO) systems. Wireless systems with massive numbers of transmit antennas (Massive MIMO) are attractive to increase the channel capacity and/or energy efficiency. However, the downlink channel state information (CSI) estimation sequence suffers increased loss in terms of lower MAC efficiency and calculation complexity. This paper proposes a CSI feedback scheme based random beamforming weights for massive MIMO when the number of users is small. In the proposed scheme, the CSI estimation beams are randomly determined and the number of beams for CSI estimation is limited. By iteratively revising the random beamforming weight, the access point can obtain the CSI of the signal space of all transmit antennas. Simulation results show that the achievable bit rate of the proposed method approaches that of the full antenna CSI.